牙齿图像分割算法研究

Study on Tooth Image Segmentation Algorithm

摘要: 牙齿图像存在边界模糊、对比度不佳的情况，传统的图像分割方法无法实现精确分割。本文提出了一种基于开闭重建和RSF & LoG模型相结合的算法用于牙齿图像分割处理，首先用开闭重建使得图像区域内部灰度趋于一致，消除金属伪影等因素的干扰，然后采用基于区域的水平集方法对图像进行分割，为克服区域内部灰度变化对水平集分割效果的干扰、以及水平集对初始设置敏感的问题。本文采用区域可调整拟合RSF模型来对图像进行分割，在RSF能量函数中增加了优化LoG的泛函能量函数以更好平滑同质区域，增强牙齿图像的边缘。实验结果表明，该算法分割效率高，鲁棒性好。

Abstract: Dental images usually have blurred boundaries and poor contrast, and traditional image segmentation methods fail to achieve accurate segmentation. This paper presents an algorithm combining open and closed reconstruction and RSF & LoG model for tooth image segmentation processing. First, the open and closed reconstruction is used to reconcile the gray scale within the image area, eliminating the interference of metal artifacts. Then, the region-based horizontal set method was used to overcome the interference of the image segmentation effect and the problems if the level set is sensitive to the initial setting. The region-adjustable fitting RSF model is used to segment the image, adding the functional energy function of the optimized LoG to the RSF energy function to better smooth the homogeneous region and enhance the edge of the tooth image. Experimental results show that the algorithm is efficient and robust.